RESEARCH / BLOG

Research,
Explained.

Read the research question, follow the authors’ methods and explore the evidence behind the article.

Research Blog

Research methodology · Research insights

How to Build a Research Novelty Matrix for Your Thesis

Build a research novelty matrix with source-linked comparisons, clear evidence labels and a testable research gap for your thesis, manuscript or PhD proposal.

By NPT Research7 min readUpdated 22 Nov 2026
The novelty toolkit03 / 09
Make the research gap visible.Compare what is known, unknown and still to be tested.
NPT Research

The Novelty Toolkit · Series 1 · Part 3 of 9

For a PhD proposal, thesis or research manuscript, an evidence-linked literature review matrix makes the research-gap argument inspectable. First find the closest prior research; then compare what each study demonstrated, what it did not establish and what your own study would need to test.

A scholar has read 35 papers and highlighted hundreds of sentences. Yet when the supervisor asks, “What exactly will your study add?”, the answer becomes vague. The problem is often not too little reading. It is that the evidence has not been compared in a usable structure.

A research novelty matrix is a side-by-side comparison of the closest prior studies and your proposed contribution. Think of it as a map of where earlier work overlaps, differs and leaves questions unanswered. Its purpose is not to decorate a proposal; it is to test whether the proposed contribution survives direct comparison.

The novelty matrixCompare evidence. Make uncertainty visible.A planned result belongs in a different category from a demonstrated one.
Studies A and B are fictional teaching examples.
DimensionStudy AStudy BProposal
Offline operationTestedUntestedPlanned
Drift correctionUnknownTestedPlanned
UncertaintyUnknownPartialPlanned
Field validationShort trialLab onlyPlanned
Unknown means verify the full text. Planned means a proposed test, not an achieved result.
Figure 1. Fictional studies illustrate how a novelty matrix distinguishes tested, unknown and planned evidence. “Planned” is an intention, not a result.

The simplest analogy: compare smartphones

Imagine three smartphone models. A glossy brochure says each is “innovative.” A comparison chart reveals which has offline translation, battery endurance or an accessible interface. But the chart still cannot tell you which phone is best until you know what users need and whether the features work. A novelty matrix has the same two stages: compare features, then evaluate their scientific importance.

Choose the comparison dimensions before filling cells

First identify the contribution claim. For our hypothetical case—adaptive sensor-drift correction for offline water-quality monitoring—use dimensions that could genuinely differentiate existing studies: the research problem, algorithmic mechanism, communication constraints, validation duration, reference standard, baselines and uncertainty reporting.

Avoid irrelevant rows such as authors’ countries, generic AI buzzwords or journal impact factors. These do not establish whether the scientific contribution is new.

Build your novelty matrix from the closest studies

For each article, record a stable reference, the precise relevant claim, the page or section containing evidence, and whether each dimension is demonstrated, partially addressed, not reported or genuinely absent. “Not mentioned in the abstract” is not the same as “not present in the paper.”

DimensionStudy A*Study B*Proposal
Primary taskWater sensingDrift compensationOffline reliable sensing
Offline operationDemonstratedNot testedPlanned
Adaptive drift modelNot reportedDemonstratedPlanned
Uncertainty estimateNot reportedPartially testedPlanned
Long-term validationShort trialLab trialField protocol proposed

*Studies A and B are fictional teaching examples. “Planned” indicates an intention, not an achieved result. In a real matrix, replace every entry with an evidence-backed statement and a source locator.

Color is useful; provenance is essential

Use text labels such as “demonstrated,” “partially studied,” “not reported” and “not tested,” with colour as a secondary cue. A missing report does not prove that a feature is absent. Keep source notes so a reviewer can verify each judgment, and make the table understandable in grayscale.

The evidence rule. Every important cell should be traceable to a page, figure, equation, supplement, patent claim or dataset. When evidence is uncertain, write “unclear—verify” rather than treating silence as proof of absence.

Keep the labels honest. “Not reported” records a reporting limit; “not tested” records an evaluation limit; “absent” requires evidence about the method itself. Reserve a separate “planned” label for your proposed contribution. Colour should reinforce these distinctions, never replace them.

The four layers of a strong novelty matrix

  1. Descriptive layer. What was actually done in each prior study?
  2. Comparative layer. Where are methods, mechanisms, populations or conditions equivalent or different?
  3. Critical layer. Which limitations are meaningful, not merely missing checkboxes?
  4. Prospective layer. Which specific new evidence will your project produce, and what could falsify the claim?

Many weak matrices stop after the second layer: they show a feature combination apparently missing from previous work. Strong matrices reach the fourth layer and explain how the study will establish the value of the difference.

Turn the matrix into a research-gap statement

From a gap to a contributionA missing feature needs a reason and a test.Build a clear chain of reasoning from prior knowledge to evidence.
  1. 01
    What is known?Describe what the nearest studies established.
  2. 02
    What remains unresolved?Identify the important unanswered question.
  3. 03
    What will you propose?State the mechanism or explanation you will test.
  4. 04
    What evidence will decide?Choose the baseline and measurements for a fair comparison.
Figure 2. A defensible research-gap statement connects established knowledge, an important unresolved issue, a proposed contribution and a test.

A useful paragraph follows this logic: “Prior studies A and B established [known result]. However, under [defined conditions], they did not establish [important unresolved issue]. We therefore propose [method or explanation], and will test it using [measurements, baseline and validation design].”

For our fictional example: “Prior work supports the feasibility of low-power water monitoring and drift correction separately, but the literature we inspected did not establish whether adaptive correction remains reliable during long offline deployments. We therefore propose a comparative field study with calibrated references, predefined error metrics and uncertainty estimates.” This is a candidate claim pending verification, not a finding about the published field.

The distinction between novelty, originality and significance helps here: the matrix describes a difference, while the surrounding argument explains its scientific value. A new combination needs a reason to be useful and a research methodology capable of testing that reason.

Add an evidence-strength column

Not all “gaps” are equally reliable. Add a note saying whether the conclusion is based on the abstract, full text, supplementary data, an exact quote or an expert’s interpretation. Mark the last checked date. This prevents a mistaken omission from becoming a bold novelty assertion.

ClaimEvidence levelAction
Study A lacks uncertainty analysisAbstract onlyRead methods + supplementary files
Study B tested only laboratory conditionsFull-text verifiedRecord section and test conditions
Proposed integration not foundMultiple databases; partial coverageRefresh preprints and citation chain

Three matrix mistakes that mislead reviewers

Cherry-picking: comparing only with weak papers while ignoring the strongest competitor. Remedy: include the nearest method even if it threatens the proposal.

Checkbox novelty: calling a study novel because no previous paper displays the same set of boxes. Remedy: explain the mechanism and its expected significance.

Future-tense inflation: treating a proposed experiment as a proven improvement. Remedy: separate “planned,” “implemented” and “validated.”

A reusable fill-in worksheet

My exact contribution claim: __________________________________________

Closest prior study and DOI: ___________________________________________

What it already demonstrates: __________________________________________

What remains unresolved, with page evidence: _____________________________

My proposed test and comparator: _______________________________________

What result would weaken my claim: _____________________________________

You can ask an AI assistant to draft an extraction table, but AI cannot certify the novelty of a research idea. Check every cell against the article, supplement or other primary source before using it in a thesis or journal submission.

How to present the matrix in a thesis or paper

Keep the published version compact: five to eight high-value dimensions and a small set of truly relevant studies are often easier to understand than a giant grid. Put detailed extraction notes, screening criteria and source passages in an appendix or supplementary file. The narrative should explain the matrix rather than simply repeat each cell.

Quick answers to common questions

Does an empty cell prove novelty?

No. It may reflect missing reporting, incomplete search or a different term. Verify the full source and related records.

Can I rank my novelty with a score?

You may use an internal rubric to prioritize questions, but do not present arbitrary weighted totals as validated novelty measurements.

Should my proposed study always have the most green cells?

No. More features do not automatically mean a better scientific contribution. A focused, carefully validated difference may be stronger than feature accumulation.

Your matrix deliverable. Produce one evidence-linked matrix, one paragraph stating the gap, and one experimental or analytical plan that could verify the contribution. If the matrix cannot support those outputs, revise the question before claiming novelty.

The matrix is an editorial comparison tool. PRISMA-S supports transparent search reporting; it does not validate this worksheet as a scoring instrument. The supplied further reading on bibliometric novelty and unusual combinations of prior work provides context rather than proof that a particular proposal is new. Patent novelty uses its own claim-based assessment.

Continue the research novelty series

Series 1: originality, evidence and scientific judgment

Begin with what makes research novel, then find the closest prior research and build a novelty matrix. Before using an assistant, examine the limits of AI novelty judgments and why the researcher owns the question. Develop the conversation with scientific prompting and six-round brainstorming, then challenge the idea against evidence. The complete worked research dialogue connects those decisions.

Series 2: from a problem to a defensible proposal

First separate a research gap from a research problem and identify a genuine problem worth investigating. Next search published literature for the gap and align the question, objectives and hypothesis. For AI-assisted work, clarify the problem without outsourcing judgment, verify the literature search and its references and compare papers in an evidence matrix. Then stress-test the research hypothesis before following the complete observation-to-proposal workflow.

Sources and further reading

  1. Rethlefsen ML et al. PRISMA-S: reporting literature searches in systematic reviews (2021). Published 26 Jan 2021
  2. Wang J, Veugelers R, Stephan P. Bias against novelty in science. Research Policy (2017).
  3. Uzzi B et al. Atypical Combinations and Scientific Impact. Science (2013). Published 25 Oct 2013
  4. WIPO. PCT International Search and Preliminary Examination Guidelines, Chapter 12: Novelty.